Project Summary HIV-1 poses a substantial health and economic burden, with more than 30 million people currently infected worldwide. The search for an effective HIV-1 vaccine remains a top priority, and a deeper understanding of how the immune system recognizes HIV-1 can help inform vaccine design. Lately, much effort has focused on understanding the antibody responses to HIV-1 infection. However, the polyclonal neutralizing antibody responses in an individual are very complex. Standard methods for mapping such responses include various experimental techniques, but more recently, computational methods were also developed. These computational methods, which we call NFP (neutralization fingerprinting), are based on analysis of serum neutralization data that is typically obtained in the very first stages of donor sample characterization, and are therefore an efficient technology for accurately mapping antibody specificities in polyclonal responses. The NFP algorithms have already become an important tool in the HIV field and are being used extensively by laboratories throughout the world, including Duke CHAVI-ID, CAPRISA, NIH VRC, and MHRP. Here, we propose to develop next-generation NFP algorithms and apply them to address biological questions with important implications for understanding the interactions between HIV-neutralizing antibodies and the virus. Specifically, we will develop and apply novel algorithms for: (1) Antibody specificity prediction with significantly improved accuracy and reliability. These algorithms will immensely improve the utility of the NFP approach for prospective identification of antibody specificities in polyclonal sera. (2) Mapping broadly neutralizing antibody responses against novel epitopes on HIV-1 Env. We will use epitope- structural analysis and computational search algorithms to identify novel Env epitopes, and will screen donor samples for the presence of related NFP signals. Promising signals for novel antibody specificities will be characterized further through collaborations. (3) Population-level analysis of broadly neutralizing antibody responses to HIV-1. We will analyze large collections of samples from diverse HIV infection cohorts in order to determine common antibody specificities elicited in response to HIV-1, as well as patterns of potential association between features of the infecting virus sequence and the elicited epitope specificities. The proposed NFP algorithms will be made available to the public, and will be useful in a number of high-impact areas in the HIV field, including mapping of antibody specificities in previously uncharacterized samples, identification of novel Env epitopes, and large-scale analysis of broadly neutralizing antibody responses within a cohort, or at a population level. Overall, this work will lead to a better understanding of the neutralizing antibody responses against HIV-1 and will build a more complete picture of the epitopes on Env. The proposed algorithmic framework should be generalizable to other important viruses, such as influenza and hepatitis C, and therefore has the potential for a far-reaching impact on public health.